Week 11: File Design and Shannon & Weaver

Reading Questions, by Zach Tomaszewski

for CIS 702, Fall 2004, taught by Dr. Martha Crosby


1. Based on your reading of "File Analysis and Design Notes", briefly define the following terms.
Chapter 1: block, (logical) record
Chapter 2: drive, disk, track, cylinder, sector, zone, seek time, latency time, read-write time.
Chapter 3: B-tree, B+-tree.
Chapter "5": Hash (or hashed file)

block
A physical record on a disk, a block is a convenient size of information to read in from the disk. Blocks may contain many records, one record, or only part of a record.
record
A logical record is a natural division in the information of a file. Example: each entry in an address book.
drive
The device that stores information on a computer. Usually includes a disk and the hardware needed to read and write to that disk (such as read-write heads, controller, etc.)
disk
A round plate or sheet covered in a magnetic coating that can maintain tiny magnetic orientations or "charges"
track
Information recorded in a concentric ring around a disk. Tracks are traced by the read-write heads of the drive.
cylinder
When a number of disks are stacked on top of each other, each read-write head can read a track on one of the disk surfaces without seeking. These tracks together are called a cylinder.
sector
Information containing sections of a track. Sectors are generally 512 bytes.
zone
On modern disks, to more efficiently use the disk surface, parts of the disk near the outside have more sectors per track than those on the inside. All the tracks with the same number of sectors per track form a zone.
seek time
The time it takes for a read-write head to move to a certain track.
latency time
The time it takes, once a head has reached a track, to access a sector in that track. (The disk has to rotate the sector around to the head.)
read-write time
The time it takes to read or write data to a sector.
B-tree
A type of binary sorted tree in which each node contains records and pointers to further nodes. Each node usually contains one more pointer than the number of records.
B+-tree
Like a B-tree, except that only leaf nodes contain records. Internal nodes contain pointers and keys.
hash
A data structure that allows for direct access for insertions and deletions. An element is inserted into a particular storage "bucket" depending on the value of its key.

2. Weaver states of the effectiveness level of communication: "It may seem at first glance undesirably narrow to imply that the purpose of all communication is to influence the conduct of the receiver. But with any reasonably broad definition of conduct, it is clear that communication either affects conduct or is without any discernible and probable affect at all" (p. 5). Do you agree or disagree? (For example: Do you agree that all communication aims to influence the conduct of the receiver? Do we ever communicate for other reasons? How do we determine whether our communication is effective?)

I disagree that we always communicate with the express purpose of affecting the conduct of the receiver. True, it is often the case that we communicate in order to reach some compromise or goal or resolution with the receiver. We may be trying to teach, bargain, beg, or cajole--all of which aim to change the conduct of the receiver.

However, other times we may communicate simply to share an idea or an emotion--simply to be heard and understood. We might tell a significant other about our day. We might publish a personal website without any particular desire to affect any particular audience, but simply to share experiences or topics of interest with whoever might wander by.

Yet Weaver does have a point when it comes to feedback about our communication. In order to determine whether the receiver is actually understanding our message, we must watch her behavior. For example, in conversation we need some sort of feedback that the receiver can hear us, understand our language, and comprehend our complete message. We at least want to see the receiver is looking at us or otherwise obviously listening. We expect a nod or comment once in a while to know that receiver is still paying attention. And we often hope to get comments or action based upon our messages. If we consider these responses as part of the receiver's conduct, it seems that we are affecting their conduct through our communication.

But where I disagree with Weaver is that this feedback--these receiver responses--is not the purpose or goal of communication. They are simply a measurable tool to determine how effectively we are communicating. It is true that we desire to communicate effectively. But I think it is important to remember the distinction between effective communication and the feedback signs that signify that effectiveness. We do not communicate solely to produce the feedback signs themselves! We may well be reaching our goal of effective communication even when those signs are not present.

3. On page 4, Shannon and Weaver outline three levels (A, B and C) of communication that are discussed throughout the chapter. They describe Level A as the technical problem, Level B as the semantic problem, and Level C as the effectiveness problem. How do these three levels compare to the three outlined by Cherry in Chapter 6 (syntax, semantics and pragmatics)?

Shannon and Weaver's Level A concerns how accurately signals can be transmitted. Level B is how precisely the transmitted symbols convey the desired meaning. And Level C is how effectively the received meaning affects conduct as desired by the sender.

Cherry's pragmatics level includes "all personal, psychological factors which distinguish one communication event from another, all questions of purpose, practical results, and value to sign users" (p. 221). Semantics intends to abstract all specifics away and find only how signs signify their designata. Syntax abstracts even further, and looks to see what "meaning", sense or redundancy is carried in the ordering of the signs themselves, regardless of their particular meanings.

On the surface, it seems like these two different theories are basically equivalent. However, I do not think they really are. Level A is concerned with transmission of signals, whatever those signals might be. The syntax level, on the other hand, aims to determine what rules guide message construction. These rules encode some very abstract meaning into the message based on its structural form. This is more along the lines of Shannon and Weaver's Level B, which concerns meaning.

Level B concerns how symbols convey meaning. This is most closely equivalent to the Semantics level, but also (as just mentioned) to the Syntax level. However, the Pragmatics level also concerns itself somewhat with meaning--the meaning of a specific message within its particular context.

Level C concerns effectiveness of the meanings conveyed to impact the receiver's conduct. Cherry's theory doesn't concern itself so much with effectiveness per se, though Pragmatics does include "practical results" and "purpose." Cherry's theory seems more open to other purposes for communication besides conduct-change.

So, in conclusion, we see that these two theories are not very compatible. Shannon and Weaver are using a simple progressive-levels view: signals are transmitted, those signals inherently carry meaning with them, and those meanings will affect conduct. Cherry is abstracting from a specific conversation and its context, to what symbols designate in the abstract for most people, to how those symbols are combined to form messages. Cherry seems concerned with meaning on all his levels, while Shannon and Weaver confine problems of meaning only to Level B.


Peterson, W., and A. Lew. "File Analysis and Design Notes" Department of ICS, University of Hawaii.

Shannon, Claude E. and Warren Weaver. The Mathematical Theory of Communication. Chicago?: University of Illinois Press, 19??.

Cherry, Colin. On human communication: a review, a survey, and a criticism. Cambridge: Technology Press of Massachusetts Institute of Technology, 1957.